examples of speech may not necessarily fall under the restrictions set out in the ICCPR, the development of the Internet and the use of social media provide a platform that can amplify this kind of speech. Attempts are underway to predict the likelihood of certain speech catalysing real world violence in certain situations. The Dangerous Speech Framework aims to find patterns in speech common before an outbreak of violence by identifying speech which targets members of a particular group, and which may have the potential to catalyse an outbreak of violence by encouraging people to condone or take part in violent acts. The framework’s guidelines are based on five variables 304, used to assess the potential impact of a particular speech. The most dangerous speech would be one for which all five variables are present: A powerful speaker with a high degree of influence over the audience; The audience has grievances and fear that the speaker can cultivate; A speech act that is clearly understood as a call to violence; A social or historical context that is propitious for violence, for any of a variety of reasons, including longstanding competition between groups for resources, lack of effort to solve grievances, or previous episodes of violence;  A means of dissemination that is influential in itself, for example because it is the sole or primary source of news for the relevant audience.     Questionnaires MCRB and IHRB developed a questionnaire based on these variables, drawing on an existing questionnaire developed by Professor Susan Benesch and the Kenyan organisation Ushahidi, which ran the Umati 305 project monitoring dangerous speech before and during the Presidential elections in Kenya in 2013. A mother-tongue Burmese-speaking researcher helped develop the questionnaire 306 for the Myanmar context and conducted research on social media websites in Myanmar. The researcher was asked specifically to search for examples of what they believed to be ‘hate speech’ and apply the questions outlined in the questionnaire to them. The researcher collected 42 examples of ‘hate speech’ over a two month period (February and March 2015), which were then analysed. Reporting the Results Studying hate speech presents an ethical dilemma: re-publishing examples may perpetuate the sentiments in the message. We have taken the decision not to re-publish statements and photographs here. The presentation of results below summarises the findings. Libre de Milles Collines (RTLM) signal. Two RTLM executives were convicted of incitement to genocide in 2007. See also Koigi Wa Wamwere, Negative Ethnicity: From Bias To Genocide (2003), Seven Stories Press, New York. Pp103-105, which describes anonymous leaflets circulated in Nazi Germany propagating hatred against Jews. More recently, leaflets were circulated provoking ethnic hatred during the break up of Yugoslavia, during the violent end to apartheid in South Africa instigating violence between the Xhosas and Zulus and have also been circulated in Kenya warning certain communities to leave their homes or be killed. 304 See: http://www.dangerousspeech.org/guidelines 305 See: http://www.ihub.co.ke/umati 306 MCRB, IHRB, DIHR, “SWIA Questionnaires” (May 2015). CHAPTER 4.2: HATE SPEECH 145 4 4.2